James Snell: Paul Prescod's rules for Extreme Web Services. Quite interesting. I absolutely agree with #'s 1, 4, 5, 7, 9 and 10. (#6 is missing for some reason). I'm still trying to determine how I feel about #'s 2, 3 and 8.  Matches my feelings pretty closely.  However,  #4 is worded a bit too "motherhood and apple pie" for my tastes... I've seen a number of underspecified protocols and people all to willing to insert their one opinions as to the original intent.   2, 3, 8 somehow seem to almost presume that the only client is a browser, particularly the statement that clients should be considered entirely stateless.  There are lots of interesting topologies with smart clients and dumb servers.

Note also that "RDF is the standard for metadata" is slipped into #5.  Smooth, eh?  Metadata is a broad term.  I also consider XML Schema metadata.

  9:24:42 PM